Search

Purpose: The study was aimed to develop and evaluate a deep learning-based radiomics to predict the histological risk categorization of thymic epithelial tumors (TETs), which can be highly informative for patient treatment planning and prognostic assessment., Method: A total of 681 patients with TETs from three independent hospitals were included and separated into derivation cohort and external test cohort. Handcrafted and deep learning features were extracted from preoperative contrast-enhanced CT images and selected to build three radiomics signatures (radiomics signature [Rad_Sig], deep learning signature [DL_Sig] and deep learning radiomics signature [DLR_Sig]) to predict risk categorization of TETs. A deep learning-based radiomic nomogram (DLRN) was then depicted to visualize the classification evaluation. The performance of predictive models was compared using the receiver operating characteristic and decision curve analysis (DCA)., Results: Among three radiomics signatures, DLR_Sig demonstrated optimum performance with an AUC of 0.883 for the derivation cohort and 0.749 for the external test cohort. Combining DLR_Sig with age and gender, DLRN was depict and exhibited optimum performance among all radiomics models with an AUC of 0.965, accuracy of 0.911, sensitivity of 0.921 and specificity of 0.902 in the derivation cohort, and an AUC of 0.786, accuracy of 0.774, sensitivity of 0.778 and specificity of 0.771 in the external test cohort. The DCA showed that DLRN had greater clinical benefit than other radiomics signatures., Conclusions: Our study developed and validated a DLRN to accurately predict the risk categorization of TETs, which has potential to facilitate individualized treatment and improve patient prognosis evaluation., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

Purpose: To evaluate the performance of machine-learning-based computed tomography (CT) radiomic analysis to differentiate high-risk thymic epithelial tumours (TETs) from low-risk TETs according to the WHO classification., Method: This retrospective study included 155 patients with a histologic diagnosis of high-risk TET (n = 72) and low-risk TET (n = 83) who underwent unenhanced CT (UECT) and contrast-enhanced CT (CECT). The radiomic features were extracted from the UECT and CECT of each patient at the largest cross-section of the lesion. The classification performance was evaluated with a nested leave-one-out cross-validation approach combining the least absolute shrinkage and selection operator feature selection and four classifiers: generalised linear model (GLM), k-nearest neighbor (KNN), support vector machine (SVM) and random forest (RF). The receiver-operating characteristic curve (ROC) and the area under the curve (AUC) were used to evaluate the performance of the classifiers., Results: The combination of UECT and CECT radiomic features demonstrated the best performance to differentiate high-risk TETs from low-risk TETs for all four classifiers. Among these classifiers, the RF had the highest AUC of 0.87, followed by GLM (AUC = 0.86), KNN (AUC = 0.86) and SVM (AUC = 0.84)., Conclusions: Machine learning-based CT radiomic analysis allows for the differentiation of high-risk TETs and low-risk TETs with excellent performance, representing a promising tool to assist clinical decision making in patients with TETs., Competing Interests: Declaration of Competing Interest We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work, there is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled., (Copyright © 2020 Elsevier B.V. All rights reserved.)

Purpose: Distinguishing between focal atelectasis (FA) and pleural dissemination (PD) is important for determining the optimal therapeutic strategy for thymic epithelial tumors (TET). This study aimed to identify useful computed tomography (CT) features for distinguishing between these two conditions., Materials and Methods: We retrospectively analyzed preoperative CT images of 27 TET, which included 40 PD and 40 FA lesions. Two radiologists independently interpreted the pleural lesions without knowing the final diagnosis. The CT images were evaluated to assess the lesion location, size, and shape, presence of a spinous shadow and ground glass opacities (GGO) near to the lesion, and the shortest distance from the lesion to the nearest peripheral pulmonary artery (PA)., Results: FA lesions tended to occur on the dorsal side (90%, P = 0.024); have shorter major and minor axes (P < 0.001), a triangular shape (43%, P = 0.002), a spinous shadow (45%, P = 0.001) and GGO (28%, P = 0.006); and be close to a peripheral PA (P = 0.007). Ninety percent of PD lesions were located in the left thorax, and all of them were ipsilateral to the tumor (both P < 0.001). The 9 examined factors exhibited sensitivity, specificity, positive predictive, and negative predictive values of 85%, 95%, 94%, and 86%, respectively, for diagnosing FA (when ≥3 factors were present), and 90%, 48%, 63%, and 83%, respectively, for diagnosing PD (when ≥4 factors were present)., Conclusion: The site, size, and shape of a lesion; the presence of a spinous shadow/GGO; and the distance to the nearest PA are useful for distinguishing between PD and FA., (Copyright © 2018 Elsevier B.V. All rights reserved.)

Purpose: To identify the MRI features of borderline epithelial ovarian tumors (BEOTs) and to differentiate BEOTs from malignant epithelial ovarian tumors (MEOTs)., Materials and Methods: The clinical and MRI data of 89 patients with a BEOT and 109 patients with a MEOT proven by surgery and histopathology were retrospectively reviewed. MRI features, including bilaterality, size, shape, margin, cystic-solid interface, configuration, papillae or nodules, signal intensity, enhancement, presence of an ipsilateral ovary, peritoneal implants and ascites were analyzed and compared. Based on the odds ratio (OR) values, the significant risk features for BEOTs were scored as 3 (OR≈∞), 2 (5≤OR<∞) or 1 (OR<5)., Results: There were 89 BEOT patients with 113 tumors [mean size of (13±6.7)cm], with bilateral ovary involvement in 24 cases. There were 109 MEOT patients with 142 tumors [(9.3±4.2)cm] with bilateral ovary involvement in 33 cases. There were eight significant risk factors for BEOTs, including round or oval shape (OR=2.714), well-defined margins (OR=3.318), clear cystic-solid interfaces (OR=5.593), purely cystic (OR=15.206), predominantly cystic with papillae or nodules (OR=2.579), exophytic papillae or nodules (OR=5.351), branching papilla (OR≈∞) and the presence of an ipsilateral ovary (OR≈∞). Based on the scoring of the eight risk factors, a cut-off score of 3.5 yielded a differential sensitivity, specificity, and accuracy of 82%, 85% and 84%, respectively., Conclusion: In contrast to MEOTs, BEOTs frequently had the following features on MRI: round or oval, with well-defined margins and clear cystic-solid interfaces, purely cystic or predominantly cystic with papillae or nodules, branching or exophytic papillae, with the presence of an ipsilateral ovary. MRI can reveal the distinct morphological features of BEOTs and MEOTs and facilitate their discrimination., (Copyright © 2017 Elsevier B.V. All rights reserved.)

Objectives: To investigate whether high-risk thymic epithelial tumor (TET) (HTET) can be differentiated from low-risk TET (LTET) using computed tomography (CT) quantitative texture analysis., Materials and Methods: The data of 39 patients (mean age, 58.6±14.1 years) (39 unenhanced CT (UECT) and 33 contrast-enhanced CT (CECT)) who underwent thymectomy for TET were retrospectively analyzed. A region of interest was placed to include the entire TET within the slice at its maximum diameter. Texture analysis was performed for images with or without a Laplacian of Gaussian filter (with various spatial scaling factors [SSFs]). Two radiologists evaluated the visual heterogeneity of TET using a 3-point scale., Results: The mean in the unfiltered image (mean0u) and entropy in the filtered image (SSF: 6mm) (entropy6u) for UECT, and the mean in the unfiltered image (mean0c) for CECT were significant parameters for differentiating between HTET and LTET as determined by logistic regression analysis. The area under the receiver operating characteristics curve (AUC) for differentiating HTET from LTET using mean0u, entropy6u, and mean0c was 0.75, 0.76, and 0.89, respectively. And the combination of mean0u and entropy6u allowed AUC of 0.87. Entropy6u provided a higher diagnostic performance compared with visual heterogeneity analysis (p≤0.018)., Conclusion: Using CT quantitative texture analysis, HTET can be differentiated from LTET with a high diagnostic performance., (Copyright © 2017 Elsevier B.V. All rights reserved.)

Objective: To assess whether types I and II epithelial ovarian cancer (EOC) differ in CT and MRI imaging features., Methods: For this retrospective study, we enrolled 65 patients with 68 ovarian lesions that have been pathologically proven to be EOC. Of these patients, 38 cases underwent MR examinations only, 15 cases underwent CT examinations only, and 12 cases completed both examinations. The clinical information [age, CA-125, menopausal status, and Ki-67] and imaging findings were compared between two types of EOCs. The diagnostic performance of image findings were assessed by receiver-operating characteristic curve(ROC) analysis. The association between EOC type and imaging features was assessed by multivariate logistic regression analysis. The random forest approach was used to build a classifier in differential diagnosis between two types of EOCs., Results: Of the 68 EOC lesions, 24 lesions were categorized as types I and other 44 lesions as type II based on the immunohistochemical results, respectively. Patients in type I EOCs were more likely to involve menopausal women and showed lower CA-125 and Ki-67 values (Ki-67<30%) than patients in type II EOCs. The imaging characteristics of type II EOCs frequently demonstrated a solid or predominantly solid mass (38.6% vs. 12.5%, P<0.05), smaller lesions (diameter <6cm; 27.3% vs. 4.2%, P<0.05), absence of mural nodules (65.9% vs. 25.9%, P=0.001), and mild enhancement (84.1% vs. 54.2%, P<0.05) compared to type I EOCs. Combination of tumor size, morphology, mural nodule, enhancement degrees (AUC=0.808) has a higher specificity (87.50%) and positive predictive value (90.0%) than any single image finding alone in differential diagnosis between two types of EOCs. The multivariate logistic regression analysis showed that enhancement degrees(OR 0.200, P<0.05),mural nodule(OR 0.158, P<0.05) significantly influence EOC classification. Random forests model identified both as the most important discriminating variables. The diagnostic accuracy of the classifier was 73.53%., Conclusions: Differences in imaging characteristics existed between two types of EOCs. Combination of several image findings improved the preoperative diagnostic performance, which is helpful for the clinical treatment and prognosis evaluation., (Copyright © 2017 Elsevier B.V. All rights reserved.)

Purpose: To evaluate the clinical usefulness of the results obtained with (18)F-FDG PET/CT in relation to CT in the preoperative staging of patients with peritoneal carcinomatosis secondary to primary or recurrent ovarian cancer candidates to cytoreductive surgery and hyperthermic intraoperative intraperitoneal chemotherapy (HIPEC)., Material and Methods: A retrospective study comparing the results obtained with CT and (18)F-FDG PET/CT in the preoperative evaluation of a series of 59 patients was performed. In all patients the peritoneal carcinomatosis index (PCI) was determined by preoperative radiological CT and 18F-FDG PET/CT and compared with surgical PCI, which was considered as reference., Results: Of the 59 patients studied, in 55 peritoneal carcinomatosis presences were confirmed (4 patients had complete responses to neoadjuvant chemotherapy). The mean surgical, CT and 18F-FDG PET/CT PCI was 9.46±7.70, 3.69±3.96 and 2.25±1.02, respectively. In the global disease detection, CT showed a higher positive likelihood ratio (LR+) than (18)F-FDG PET/CT (15.3, 95% CI 8.35-28.04 vs. 3.47, 95% CI 3.36-5.11) and a lower negative likelihood ratio (LR-) than 18F-FDG PET/CT (0.67, 95% CI 0.61-0.73 vs. 0.82, 95% CI 0.76-0.88). In every region of the abdomen the CT showed a greater LR+ than 18F-FDG PET/CT and a lower LR- than 18F-FDG PET/CT., Conclusions: CT showed the best diagnostic results compared to (18)F-FDG PET/CT to confirme the presence of peritoneal disease. The lower performance of the (18)F-FDG PET/CT suggests that the main utility of (18)F-FDG PET/CT is to evaluate a possible metastatic extraperitoneal spread of the disease., (Copyright © 2016 Elsevier Ireland Ltd. All rights reserved.)